In Maxent model evolution the output result using AUC value; If AUC is more then 0.7 and close to 1 that result is best. But if we ruled out AUC value for evolution then what are the methods for selection the best result? Using Different variables for a single species I run Maxent model and I have four (4) output. Then how to I evaluate the best result for my study? How to I Compare those result and select a best result except AUC value?
Hi,
there are actually many options and there are certainly experts for statistical procedures that can provide more help than me.
That said, you can:
1) Use R packages that streamline variable selection and model evaluation in MaxEnt. MaxentVariableSelection and ENMtools are two prominent ones.
2) Look at variable contributions and keep to the rule of thumb: "when two or more models are statistically and biologically sound, prefer the simpler one".
2) Assess you model in-situ: either go there yourself or ask regional experts whether the real situation in the field is in accordance with the projected habitat suitability (from your model). This should NOT relpace statistical testing, you should do both (Morales, Fernández, & Baca-González 2017).
Cheers,
Florian
Morales NS, Fernández IC, Baca-González V (2017) MaxEnt’s parameter configuration and small samples: are we paying attention to recommendations? A systematic review. PeerJ 5:e3093. https://doi.org/10.7717/peerj.3093
For other test for "best" fit of SDM including maxent output and complementary considerations (optimizing transferability; minimizing multi-collinearity of independent variables) for testing and model selection see the following:
Article Transferability of species distribution models: The case of ...
AIC Akaike Information Criterion may also be relevant
Be welcome to share your findings of additional testing
You can do sensitivity testing, such as comparing variance across the models :)
The true skill statistic (TSS) is a good option, see Allouche et al. (2006) Article Assessing the accuracy of species distribution models: preva...
Alternative to AUC are TSS, kappa, sensitivity test, among others.
Hello there
see the KUENM tools and applications for the AUC from Maxent
https://github.com/marlonecobos/kuenm
Cobos, M. E., Peterson, A. T., Barve, N., & Osorio-Olvera, L. (2019). kuenm: an R package for detailed development of ecological niche models using Maxent. PeerJ, 7, e6281.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs